Drug addiction is characterized by a transition from volitional drug use to drug use that is increasingly habitual and compulsive. Thus, behavior among addicted individuals, including drug-seeking behavior, may be partly controlled by motivational processes that occur below conscious awareness. A recent review indeed suggested that such impaired insight may be a hallmark of drug addiction. Using functional magnetic resonance imaging (fMRI), the current proposal seeks to inspect the neural correlates of this putative insight deficit. Specifically, this proposal will explore functioning of the anterior insula (AI) and anterior cingulate cortex (ACC), both posited to play key roles in insight and self-awareness, in cocaine addiction. To accomplish this aim, an event- related fMRI color word Stroop task, a classical neuropsychological conflict task, will be employed, which will be modified for the purposes of this proposal to test for insight into task-related errors. This modified fMRI Stroop task will be performed by three study groups: actively using cocaine addicted individuals (CUD+), treatment- seeking cocaine addicted individuals (CUD-), and healthy control subjects (HC). This fMRI insight-into-errors data will then be used to predict subsequent cocaine-related choice behavior. This latter study component will be accomplished through the use of the Preference for Image Viewing, Probabilistic (PIV-P) task, which examines choice for cocaine-related pictures compared with choice for other pleasant, unpleasant, or neutrally valenced pictures. This design will allow for testing of whether impaired insight and its neural underpinnings predict drug- seeking behavior. In addition, the CUD- individuals, who will be recruited from local treatment facilities, will complete this same PIV-P task bi-monthly during treatment-as-usual at these same facilities. This prospective design will allow for testing of the validity of 'incubated craving' (an escalating propensity for relapse during abstinence) in human addicted individuals as they progress through treatment (seven follow-up assessments are planned), and to further establish whether the fMRI insight data predict its time course. If successful, results of this proposal may help change the prevalent perception of drug addicted individuals as displaying treatment- resistant and oppositional denial. Instead, these results may identify drug addicted individuals as struggling with a neurally-based, and possibly treatment-amenable, insight deficit.